Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00063
Ratheesh Ravindran, M. Santora, M. Faied, Mohammad Fanaei
One of the most crucial enabling technologies for automated driving systems is the ability to reliably detect and classify a wide range of traffic signs in various driving conditions at different distances. Due to the complexity and dynamic nature of driving environments, it is difficult to reliably detect traffic signs with conventional image processing methods. Artificial intelligence in combination with image processing has proven to be a great success to address this problem in recent studies. This paper focuses on the selection of Deep Neural Networks (DNN) based on the application-oriented performance by taking into consideration the mean Average Precision (mAP) and Frames Per Second (FPS) as the major evaluation criteria. Faster Region-based Convolutional Neural Network (Faster R-CNN) is a newly proposed DNN in the literature that has proven to exhibit a balanced tradeoff between mAP and FPS performance measures. This paper starts with a DNN transfer learning and then implements the Faster R-CNN algorithm for the real-time detection and classification of traffic signs using the Robot Operating System (ROS). To reduce the errors due to DNN inaccurate detection, Tesseract" is added to detect the text in the identified traffic signs. The German Traffic Sign Detection Benchmark (GTSDB) dataset is used in this paper, and additional dataset are created to solve the lack of certain traffic signs in the GTSDB dataset. Simulation with ROS-Gazebo and real-time trials using the Polaris Gem e2 equipped with NVIDIA Drive PX2 demonstrate the efficiency of the proposed integration of DNN with Tesseract in detecting and classifying a wide range of traffic signs.
{"title":"Traffic Sign Identification Using Deep Learning","authors":"Ratheesh Ravindran, M. Santora, M. Faied, Mohammad Fanaei","doi":"10.1109/CSCI49370.2019.00063","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00063","url":null,"abstract":"One of the most crucial enabling technologies for automated driving systems is the ability to reliably detect and classify a wide range of traffic signs in various driving conditions at different distances. Due to the complexity and dynamic nature of driving environments, it is difficult to reliably detect traffic signs with conventional image processing methods. Artificial intelligence in combination with image processing has proven to be a great success to address this problem in recent studies. This paper focuses on the selection of Deep Neural Networks (DNN) based on the application-oriented performance by taking into consideration the mean Average Precision (mAP) and Frames Per Second (FPS) as the major evaluation criteria. Faster Region-based Convolutional Neural Network (Faster R-CNN) is a newly proposed DNN in the literature that has proven to exhibit a balanced tradeoff between mAP and FPS performance measures. This paper starts with a DNN transfer learning and then implements the Faster R-CNN algorithm for the real-time detection and classification of traffic signs using the Robot Operating System (ROS). To reduce the errors due to DNN inaccurate detection, Tesseract\" is added to detect the text in the identified traffic signs. The German Traffic Sign Detection Benchmark (GTSDB) dataset is used in this paper, and additional dataset are created to solve the lack of certain traffic signs in the GTSDB dataset. Simulation with ROS-Gazebo and real-time trials using the Polaris Gem e2 equipped with NVIDIA Drive PX2 demonstrate the efficiency of the proposed integration of DNN with Tesseract in detecting and classifying a wide range of traffic signs.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115574183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00096
E. Espinosa-Juárez, Jorge Luis Solano-Gallegos, F. Ornelas‐Tellez
This paper presents the problem of economic dispatch for an electrical system with unconventional energy sources and energy storage. The economic dispatch is considered for demand variations over 24 hours, taking into account the forecast of solar energy for one hour ahead, based on the autoregressive process. The implemented algorithm allows analysis of economic dispatch under different restrictions. A case study is shown, where different levels of renewable energy penetration into the system are considered and the effectiveness of the implemented algorithm is observed
{"title":"Economic Dispatch for Power System with Short-Term Solar Power Forecast","authors":"E. Espinosa-Juárez, Jorge Luis Solano-Gallegos, F. Ornelas‐Tellez","doi":"10.1109/CSCI49370.2019.00096","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00096","url":null,"abstract":"This paper presents the problem of economic dispatch for an electrical system with unconventional energy sources and energy storage. The economic dispatch is considered for demand variations over 24 hours, taking into account the forecast of solar energy for one hour ahead, based on the autoregressive process. The implemented algorithm allows analysis of economic dispatch under different restrictions. A case study is shown, where different levels of renewable energy penetration into the system are considered and the effectiveness of the implemented algorithm is observed","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"177 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115585147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00147
Sahar Voghoei, Navid Hashemi Tonekaboni, D. Yazdansepas, H. Arabnia
It has been generally believed that higher participation in discussion forums in online classes would result in better student performance. To better understand this correlation on a large scale, we have studied 291 distinct online courses offered during Summer 2019 at Georgia Gwinnett College. Several studies in the literature have focused on analyzing the data from the Massive Open Online Courses (MOOCs). However, in this research, we have focused on University-based Online Courses (UOCs) for undergraduate students, where the curriculum enforces students to take these courses. Although a higher participation rate in online forums has a direct correlation with a higher grade in MOOCs, in OUCs, students with top grades are not necessarily the most active students. Our analysis shows a consistent pattern in UOCs where during the first two-thirds of the semester, students who belong to the GPA range of ~70 to ~80 percentile of the class have the highest rate of participation, while during the last one-third of the semester, the ones who belong to the GPA range of ~87 to ~93 percentile, contribute the most. On the other hand, we found out that the common characteristic of top students in all classes, is their consistency in participation throughout the semester, regardless of the number of their posts.
{"title":"University Online Courses: Correlation between Students' Participation Rate and Academic Performance","authors":"Sahar Voghoei, Navid Hashemi Tonekaboni, D. Yazdansepas, H. Arabnia","doi":"10.1109/CSCI49370.2019.00147","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00147","url":null,"abstract":"It has been generally believed that higher participation in discussion forums in online classes would result in better student performance. To better understand this correlation on a large scale, we have studied 291 distinct online courses offered during Summer 2019 at Georgia Gwinnett College. Several studies in the literature have focused on analyzing the data from the Massive Open Online Courses (MOOCs). However, in this research, we have focused on University-based Online Courses (UOCs) for undergraduate students, where the curriculum enforces students to take these courses. Although a higher participation rate in online forums has a direct correlation with a higher grade in MOOCs, in OUCs, students with top grades are not necessarily the most active students. Our analysis shows a consistent pattern in UOCs where during the first two-thirds of the semester, students who belong to the GPA range of ~70 to ~80 percentile of the class have the highest rate of participation, while during the last one-third of the semester, the ones who belong to the GPA range of ~87 to ~93 percentile, contribute the most. On the other hand, we found out that the common characteristic of top students in all classes, is their consistency in participation throughout the semester, regardless of the number of their posts.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117260599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00067
C. Tappert
This paper evaluates candidates for the father of deep learning. We conclude that Frank Rosenblatt developed and explored all the basic ingredients of the deep learning systems of today, and that he should be recognized as a Father of Deep Learning, perhaps together with Hinton, LeCun and Bengio who have just received the Turing Award as the fathers of the deep learning revolution.
{"title":"Who Is the Father of Deep Learning?","authors":"C. Tappert","doi":"10.1109/CSCI49370.2019.00067","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00067","url":null,"abstract":"This paper evaluates candidates for the father of deep learning. We conclude that Frank Rosenblatt developed and explored all the basic ingredients of the deep learning systems of today, and that he should be recognized as a Father of Deep Learning, perhaps together with Hinton, LeCun and Bengio who have just received the Turing Award as the fathers of the deep learning revolution.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115784584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00139
R. Hasanah, Rakhmat Ramadhan, H. Suyono, T. Taufik
This paper presents a comparative study on the performance of PID and Voltage Mode Control (VMC) in a step-down voltage or buck DC-DC converter. The converter is being used in a smart wall plug for powering electrical devices in future smart house or building. Computer simulations using Simulink were performed to model the controllers in the converter and to investigate their performance. Results indicate that longer time is required by the VMC to reach a similar steady state condition as that acquired by the PID on the output voltage of the converter. Additionally, the steady state error on the output voltage from the PID was observed to be less than 1%, which is better than percent error obtained from the VMC.
{"title":"Performance Study of PID and Voltage Mode Controllers in Voltage Regulator for Smart DC Wall-Plug","authors":"R. Hasanah, Rakhmat Ramadhan, H. Suyono, T. Taufik","doi":"10.1109/CSCI49370.2019.00139","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00139","url":null,"abstract":"This paper presents a comparative study on the performance of PID and Voltage Mode Control (VMC) in a step-down voltage or buck DC-DC converter. The converter is being used in a smart wall plug for powering electrical devices in future smart house or building. Computer simulations using Simulink were performed to model the controllers in the converter and to investigate their performance. Results indicate that longer time is required by the VMC to reach a similar steady state condition as that acquired by the PID on the output voltage of the converter. Additionally, the steady state error on the output voltage from the PID was observed to be less than 1%, which is better than percent error obtained from the VMC.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123430062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/csci49370.2019.00002
{"title":"[Title page iii]","authors":"","doi":"10.1109/csci49370.2019.00002","DOIUrl":"https://doi.org/10.1109/csci49370.2019.00002","url":null,"abstract":"","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"311 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123147838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00216
L. Alhazzaa, Anneliese Amschler Andrews
In any software development organization, reliability is crucial. Defect prediction is key in providing management with the tools for release planning. To predict defects we ask the question of how much data is required to make usable predictions? When testing, a rule of thumb is to start defect prediction after 60% of system test has been accomplished. In an operational phase, managers cannot usually determine what constitutes 60% of a release and might not want to wait that long to start defect prediction. Here we discuss the trade-offs between the need of early predictions versus making more accurate predictions.
{"title":"Trade-Offs between Early Software Defect Prediction versus Prediction Accuracy","authors":"L. Alhazzaa, Anneliese Amschler Andrews","doi":"10.1109/CSCI49370.2019.00216","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00216","url":null,"abstract":"In any software development organization, reliability is crucial. Defect prediction is key in providing management with the tools for release planning. To predict defects we ask the question of how much data is required to make usable predictions? When testing, a rule of thumb is to start defect prediction after 60% of system test has been accomplished. In an operational phase, managers cannot usually determine what constitutes 60% of a release and might not want to wait that long to start defect prediction. Here we discuss the trade-offs between the need of early predictions versus making more accurate predictions.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121050000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00094
Jun Zhang, Weiwei Zhu, Fangyang Shen
Integral equations come from a wide range of applications. Laplace transform has been playing an important role in mathematics; it is very powerful and widely used in solving integral equations, however, such a traditional method suffers a serious drawback, which is the calculation of inverse Laplace transform. Such a kind of inverse calculation is problematic or impossible, except some very simple functions. Sumudu transform is a new integral transform with nice features like Laplace transform, in addition, it provides new methodology for problem solving. In this work, a new computational method is proposed to solve integral equations, the new method incorporates useful features from both Laplace transform and Sumudu transform such that the calculation of the inverse Laplace transform is avoided. In addition, it is demonstrated with implementations that the new method and techniques presented in this work can be implemented in computer algebra systems such as Maple to solve Volterra convolution integral equations and mixed differential Volterra convolution integral equations automatically
{"title":"New Algorithms to Solve Integral Equations Automatically","authors":"Jun Zhang, Weiwei Zhu, Fangyang Shen","doi":"10.1109/CSCI49370.2019.00094","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00094","url":null,"abstract":"Integral equations come from a wide range of applications. Laplace transform has been playing an important role in mathematics; it is very powerful and widely used in solving integral equations, however, such a traditional method suffers a serious drawback, which is the calculation of inverse Laplace transform. Such a kind of inverse calculation is problematic or impossible, except some very simple functions. Sumudu transform is a new integral transform with nice features like Laplace transform, in addition, it provides new methodology for problem solving. In this work, a new computational method is proposed to solve integral equations, the new method incorporates useful features from both Laplace transform and Sumudu transform such that the calculation of the inverse Laplace transform is avoided. In addition, it is demonstrated with implementations that the new method and techniques presented in this work can be implemented in computer algebra systems such as Maple to solve Volterra convolution integral equations and mixed differential Volterra convolution integral equations automatically","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124286061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/csci49370.2019.00084
Seungchul Lee, Daeyoung Kim
In manufacturing industry, significant productivity losses arise due to equipment failures. Therefore, it is an important task to prevent the equipment from failure by monitoring each machine's sensor data in advance. However, most of the current developed systems have been only focused on monitoring the sensor data and have a difficulty in applying advanced algorithms to the real-time stream data. To address issues, we implemented an intelligent system that employs real-time streaming engine loaded with the machine learning libraries for predictive maintenance analysis. By applying a deep-learning based model to the real-time streaming data, we can provide not only trends of raw sensor data but also give an indicator representing an equipment's status in real-time. We anticipate that our system contributes to recognize the equipment's status by monitoring the indicator for productivity improvement in manufacturing industry in real-time.
{"title":"A Real-Time Based Intelligent System for Predicting Equipment Status","authors":"Seungchul Lee, Daeyoung Kim","doi":"10.1109/csci49370.2019.00084","DOIUrl":"https://doi.org/10.1109/csci49370.2019.00084","url":null,"abstract":"In manufacturing industry, significant productivity losses arise due to equipment failures. Therefore, it is an important task to prevent the equipment from failure by monitoring each machine's sensor data in advance. However, most of the current developed systems have been only focused on monitoring the sensor data and have a difficulty in applying advanced algorithms to the real-time stream data. To address issues, we implemented an intelligent system that employs real-time streaming engine loaded with the machine learning libraries for predictive maintenance analysis. By applying a deep-learning based model to the real-time streaming data, we can provide not only trends of raw sensor data but also give an indicator representing an equipment's status in real-time. We anticipate that our system contributes to recognize the equipment's status by monitoring the indicator for productivity improvement in manufacturing industry in real-time.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126338065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-12-01DOI: 10.1109/CSCI49370.2019.00071
Ivan Ortiz Garcés, María Cazares, R. Andrade
The number of phishing attacks has increased in Latin America, exceeding the operational skills of cybersecurity analysts. The cognitive security application proposes the use of bigdata, machine learning, and data analytics to improve response times in attack detection. This paper presents an investigation about the analysis of anomalous behavior related with phishing web attacks and how machine learning techniques can be an option to face the problem. This analysis is made with the use of an contaminated data sets, and python tools for developing machine learning for detect phishing attacks through of the analysis of URLs to determinate if are good or bad URLs in base of specific characteristics of the URLs, with the goal of provide realtime information for take proactive decisions that minimize the impact of an attack.
{"title":"Detection of Phishing Attacks with Machine Learning Techniques in Cognitive Security Architecture","authors":"Ivan Ortiz Garcés, María Cazares, R. Andrade","doi":"10.1109/CSCI49370.2019.00071","DOIUrl":"https://doi.org/10.1109/CSCI49370.2019.00071","url":null,"abstract":"The number of phishing attacks has increased in Latin America, exceeding the operational skills of cybersecurity analysts. The cognitive security application proposes the use of bigdata, machine learning, and data analytics to improve response times in attack detection. This paper presents an investigation about the analysis of anomalous behavior related with phishing web attacks and how machine learning techniques can be an option to face the problem. This analysis is made with the use of an contaminated data sets, and python tools for developing machine learning for detect phishing attacks through of the analysis of URLs to determinate if are good or bad URLs in base of specific characteristics of the URLs, with the goal of provide realtime information for take proactive decisions that minimize the impact of an attack.","PeriodicalId":103662,"journal":{"name":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125904851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：